An echo, or acoustic reflection, occurs when an acoustic wave encounters an object such as an enclosure wall. When a reflection occurs, the reflected wave interacts with the wave that was originally directed towards the object causing the reflection. The waves are often labeled as the incident and reflected waves. At low amplitudes the two waves interact in simple superposition, adding to produce a sound pressure pattern in space. In a typical system, the acoustic wave/reflection result occurs in three dimensions. In an environment with walls that reflect most of the wave directed at them, points can be seen where the resultant sound pressure decreases to 10 percent or less of the amplitude of the initial incident wave.
The addition of incident and reflected waves produce a sound pressure pattern that is typically quite complicated. This pattern is also dependent on the frequencies of the waves. A complex waveform containing many frequencies will have a set of reflection patterns, each dependent on an individual frequency. The result is that it is very difficult to know the sound pressure at any point in a 3 dimensional space that contains reflective surfaces.
A device to be tested, be it a sound emission device like a speaker, a sound reception device like a microphone, or a combination device like a hearing aid, has apparent acoustic properties affected by the environment in which it is tested. If the environment contains surfaces that reflect acoustic waves, the properties of the device under test are subject to reflection artifacts. Unfortunately, surfaces and objects reflect acoustic waves. The best that can be done is to provide a surface, or combination of surfaces, that have small acoustic reflections that do not significantly affect the measurement of the device under test.
Some acoustic devices are constructed to have directional properties. For these devices it is important to measure device characteristics in an acoustic environment with few reflections. Often a chamber known as an “anechoic chamber” is used for such testing. As noted above, there is no such thing as a chamber that has no reflections. However, chambers have been constructed that have sufficient attenuation of reflections to allow reasonable testing of these directional devices. Typically, these chambers are large. Current technology uses sound absorbing wedges that are a substantial percentage of a wavelength deep. For low frequency operation, the chamber must be large in order so that the walls formed by the wedges are thick enough to absorb the sound waves.
The wedges are typically constructed using a wire form that is stuffed with fiberglass. The wire itself reflects a certain amount of acoustic energy, as does the fiberglass. If the wedges have relatively sharp edges, only very high frequencies will be reflected off of the wedge edges, and only a small percentage of the waves will be reflected back toward the generator of the incident wave.
The wedges are also constructed with relatively sharp angles. Waves that encounter a wedge side surface will reflect off the surface. The sharp angles of the wedge sides cause the wave reflection to move inward toward a surface of another adjacent wedge. The adjacent wedge then reflects the wave back toward a deeper portion of the first wedge. Thus, the acoustic wave works its way towards the wedge base and hopefully is mostly absorbed by the time the wave reaches the wedge base. Of course, the wedges hold fiberglass, which is a good absorber of sound. Therefore most of the signal that hits the side of the wedge is absorbed in the fiberglass material and only a small percentage is reflected.
The reflection behavior of a wave from a surface is dependent on the dimensions of the surface and the wavelength. If a sound chamber is small compared with the wavelength, then reflections may be ignored and the enclosure may be thought of as a pressure box. Relatively small anechoic chambers are therefore not effective for low frequencies with wavelengths that exceed the dimensions of the chamber. The damping action of the wedges in a sound chamber is also reduced when the dimensions of the wedges are an appreciable percentage of a wavelength.
In recent years, certain types of open cell foams have been available for acoustic damping of surfaces in chambers and rooms. Some of these foams have desirable properties that reduce sound transmission through the foam and also attenuate reflections of waves directed at the surface of the foam. The foams come in a variety of densities and construction.
As with fiberglass, sound incident on a foam surface is partially reflected as well as attenuated upon entering the material. A portion of a sound wave hitting a simple surface covered with a thickness of foam will be reflected from the surface of the foam and a portion will travel into the foam. If the thickness of the foam is increased, sound will be attenuated as it proceeds through the foam. When the sound travels completely through the foam thickness, it will eventually encounter the underlying surface. For example, a concrete or wood wall surface that supports the foam. Most of the sound encountering this surface will be reflected back into the foam material and undergo further attenuation before emerging from its outer surface.
Thus an incident sound wave encountering a simple plane damping surface will split. Some will be reflected and the rest will travel into the damping material and eventually emerge attenuated in amplitude. This returning attenuated sound will add to the initially reflected sound from the front surface of the damping material. The portion of the incident sound that is initially reflected from the front surface appears to be unaffected by an increase in the thickness of the damping material.
Acoustic devices of all types, including receivers (microphones) and generators (speakers), have a pattern to the way they operate. The sound that they receive or generate typically has a 3 dimensional directional component. For speakers, the sound emanating from the device is typically directed in one particular direction more than other directions. The same is sometimes true for microphones. Sometimes microphones or devices that employ microphones are constructed in a way that enhances the directional capability of the device. The directional characteristic of the acoustic device is also typically dependent on the acoustic frequency. Because of the wavelength nature of a sound wave, devices handle different frequencies in different ways.
From an engineering and manufacturing perspective, it is desirable to know the pattern that the acoustic device exhibits at each frequency. Tests are typically run on the device in areas that are as free of reflected sound as possible, such as in an anechoic chamber or in a chamber free of echo. Sounds from speakers can then be tested for their directional pattern. Microphones can be located at different points in the sound generation path of the speaker to collect this information. Or the microphone can be kept in one spot and the speaker moved to different orientations for the test.
Directional microphones can be tested in similar ways. The microphone can be held in a constant position and the sound source moved to make a test, or the microphone orientation can be changed, holding a fixed sound source location.
The typical system will test the speaker or microphone directional pattern characteristic one frequency at a time. The data is often displayed in a graphical format called a polar plot. The plot exhibits the directional performance of the device for that frequency in a particular plane of operation and is labeled as amplitude vs. angular position within that plane.
Another possible display of the information is in the form of a series of overlaid frequency response curves. Each curve has a different positional angle from a reference angle. Sometimes this information will be confined to the angle at which the greatest sensitivity or efficiency is demonstrated and the angle at which the sound is at the lowest amplitude. There are a number of ways in which the information may be displayed.